libact.models.multilabel package

Submodules

libact.models.multilabel.binary_relevance module

This module contains implementation of binary relevance for multi-label classification problems

class libact.models.multilabel.binary_relevance.BinaryRelevance(base_clf, n_jobs=1)

Bases: libact.base.interfaces.MultilabelModel

Binary Relevance

base_clf : libact.models object instances
If wanting to use predict_proba, base_clf are required to support predict_proba method.
n_jobs : int, optional, default: 1
The number of jobs to use for the computation. If -1 all CPUs are used. If 1 is given, no parallel computing code is used at all, which is useful for debugging. For n_jobs below -1, (n_cpus + 1 + n_jobs) are used. Thus for n_jobs = -2, all CPUs but one are used.

References

[1]Tsoumakas, Grigorios, Ioannis Katakis, and Ioannis Vlahavas. “Mining multi-label data.” Data mining and knowledge discovery handbook. Springer US, 2009. 667-685.
predict(X)

Predict labels.

Parameters:X (array-like, shape=(n_samples, n_features)) – Feature vector.
Returns:pred – Predicted labels of given feature vector.
Return type:numpy array, shape=(n_samples, n_labels)
predict_proba(X)

Predict the probability of being 1 for each label.

Parameters:X (array-like, shape=(n_samples, n_features)) – Feature vector.
Returns:pred – Predicted probability of each label.
Return type:numpy array, shape=(n_samples, n_labels)
predict_real(X)

Predict the probability of being 1 for each label.

Parameters:X (array-like, shape=(n_samples, n_features)) – Feature vector.
Returns:pred – Predicted probability of each label.
Return type:numpy array, shape=(n_samples, n_labels)
score(testing_dataset, criterion='hamming')

Return the mean accuracy on the test dataset

Parameters:
  • testing_dataset (Dataset object) – The testing dataset used to measure the perforance of the trained model.
  • criterion (['hamming', 'f1']) – instance-wise criterion.
Returns:

score – Mean accuracy of self.predict(X) wrt. y.
Return type:

float
train(dataset)

Train model with given feature.

Parameters:
  • X (array-like, shape=(n_samples, n_features)) – Train feature vector.
  • Y (array-like, shape=(n_samples, n_labels)) – Target labels.
clfs_

list of libact.models object instances – Classifier instances.

Returns:self – Retuen self.
Return type:object

libact.models.multilabel.dummy_clf module

This module provides a dummy classifier, since in multi-label active learning problem, it is common to see label being all zero in training set. We will let this classifier handles this condition.

class libact.models.multilabel.dummy_clf.DummyClf

Bases: object

This classifier handles training sets with only 0s or 1s to unify the interface.

fit(X, y)
predict(X)
predict_proba(X)
predict_real(X)
train(dataset)

Module contents

Concrete model classes.